RNA from LPS-stirnulated macrophages induces the release of tumour necrosis factor-α and interleukin-1 by resident macrophages

The effect of exogenous RNA on many cellular functions has been studied in a variety of eukaryotic cells but there are few reports on macrophages. In the present study, it is demonstrated that cytoplasmatic RNA extracted from rat macrophages stimulated with Escherichia coli lipopolysaccharide (LPS), referred to as L-RNA, induced the release of TNF-α and IL-1 from monolayers of peritoneal resident macrophages. The activity of L-RNA was not altered by polymyxin B but was abolished by ribonuclease (RNase) pretreatment, indicating the absence of LPS contamination and that the integrity of the polynucleotide chain is essential for this activity. Both the poly A(−) and poly A(+) fractions obtained from L-RNA applied to oligo(dT)–cellulose chromatography induced TNF-α and IL-1 release. The L-RNA-induced cytokine release was inhibited by dexamethasone and seemed to be dependent on protein synthesis since this effect was abolished by cycloheximide or actinomycin-D. The LPS-stimulated macrophages, when pre-incubated with [5-3H]-uridine, secreted a trichloroacetic acid (TCA) precipitable material which was sensitive to RNase and KOH hydrolysis, suggesting that the material is RNA. This substance was also released from macrophage monolayers stimulated with IL-1β but not with TNF-α, IL-6 or IL-8. The substance secreted (3H-RNA) sediments in the 4–5S region of a 5–20% sucrose gradient. These results show that L-RNA induces cytokine secretion by macrophage monolayers and support the idea that, during inflammation, stimulated macrophages could release RNA which may further induce the release of cytokines by the resident cell population.


Introduction
There is considerable evidence indicating that exogenously supplied RNA molecules can be incorporated into eukaryotic cells and can exert a variety of biological effects both in vitro and in vivo. [1][2][3][4][5] Cell differentiation, genetic transformation and immunological effects induced by exogenous RNA have been investigated over the past few years. For example, RNA isolated from the endoderm of normal embryos of the amphibian A mbystoma mexicanum causes the differentiation of mutant heart tissue into functional, contractile cardiac muscle. 6 The genetic nutritional deficiencies of a strain ofA spergillus nidulans have been corrected by genetic transformation using RNA extracted from the wild strain of this fungus. Recently, the authors demonstrated the transfer, to human lymphocytes, of cellular immunoreactivity to a highly conserved and immunodominant, synthetic epitope of the glycoprotein 41 of HIV-1, by using RNA isolated from animals immunized with this synthetic peptide. 993 Rapid Communications of Oxford Ltd The effect of exogenous RNA has also been detected in vivo. The direct injection of purified mRNA for the bacterial enzyme chloramphenicol acetyl transferase (CAT) into mouse skeletal muscle results in significant CAT activity in these muscle cells. Similarly, the injection of synthetic mRNA for the Wnt oncogene family induces cellular differentiation in Xenopus embryos, associated with the formation of specific embryonic structures. 1 The RNA obtained from the lymphoid organs of animals immunized with carcinoma antigens from human colon induces memory T-lymphocytes against these tumour antigens in vivo. 1 It is now becoming clear that the cytokines constitute a link between cellular injury or recognition of non-self and the development of the inflammatory and/or immune response. Among the cytokines, interleukin 1 (IL-1) and tumour necrosis factor-o (TNF-o) play an important role in triggering and in the maintenance of these events. 12 '13 Although several cells release cytokines, the resident macrophages can be considered the R. A. Ribeiro et al. most active cytokine secreting cells. 14'1s When activated by exogenous and/or endogenous stimuli, these macrophages release various cytokines, including TNF-, IL-1, IL-6 and IL-8, which induce the local and systemic events of the defence response. 16,17 Recently, it has been suggested that RNA may function as an intercellular mediator, acting over short distances, leading to the possibility of the transfer of biological information from one cell to another via RNA molecules. 18 In the present study, we investigated whether RNA extracted from LPS-activated macrophages could stimulate the release of TNFand IL-1 from naive macrophages in culture. We also investigated whether macrophages stimulated by endotoxin released RNA into the culture medium. The results of the present investigation are consistent with the idea that RNA may be involved as an intercellular modulator of cytokine release from macrophages.

Materials and Methods
Preparation of macrophage monolayers: Rat peritoneal cavities were stimulated by the injection of a 3% thioglycollate solution (10ml/animal). Four days later, the peritoneal macrophages were harvested in RPMI medium and allowed to adhere to plastic tissue culture dishes for 1 h at 37C, in an atmosphere of air containing 5% CO2. The monolayers were then washed three times with phosphate buffered saline (PBS), pH 7.4, and incubated for 1 h at 37C in RPMI medium containing 10/,g/ml of Escherichia coli lipopolysaccharide (LPS). The supernatants were discarded and the monolayers washed a further three times with PBS. The cells were harvested with a disposable rubber policeman and were used for RNA extraction after centrifugation (1000 x g for 10 rain). RNA extraction: Cytoplasmic RNA was extracted from LPS-stimulated macrophages using the cold phenol method described by White and De Lucca. 19 The macrophage pellet was homogenized in Tris-HC1 buffer, pH 9.0, containing 0.01 M KC1 and 0.25 M sucrose (2.5 ml/g macrophage pellet, wet weight) and centrifuged at 12 000 x g, for 15 min at 4C. The supernatant was diluted two-fold with the homogenization buffer and used as the source of cytoplasmic RNA.
The supernatant was deproteinized by the addition of an equal volume of 88% (v/v) phenol containing 0.05% bentonite. The mixture was shaken for 10 min and then centrifuged at 2 000 x g for 10 min at 4C. The  The concentration of RNA was calculated by measuring absorbance at 260 nm (1 absorbance unit 40/zg of RNA/ml). Only those preparations of RNA with a ratio A260/a280 of about 1.9-2.0 were used. The RNA preparations obtained from LPS-stimulated macrophages are referred to as L-RNA. Samples of cytoplasmic RNA obtained from non-stimulated macrophage monolayers by the same extraction procedure are referred to as normal RNA (N-RNA).
The integrity of the RNA was routinely evaluated using SDS-polyacrylamide gel electrophoresis according to Bertolini and De Lucca. 2 The gels were fixed in acetic acid and scanned for absorbance at 260nm in a spectrophotometer equipped with a linear transport.
Oligo(dT)-cellulose chromatography: The poly(A)-containing RNA (poly A(4-)-RNA) and the poly(a)lacking RNA (poly A(--)-RNA) fractions were separated from the bulk of the cytoplasmic RNA (total RNA) by affinity chromatography on an oligo(dT)-cellulose column as described by Aviv and Leder. 21 Enzymatic treatment of L-RNA: The L-RNA preparations were subjected to enzymatic treatment using a mixture containing ribonuclease T (EC 3.1.27.3; guanyloRNase) and bovine pancreatic ribonuclease (EC 3.1.27.5; RNase type I-A). The enzymes were diluted in 0.85% NaC1 in the ratio of 1/,g of each RNase to 10/.zg L-RNA and the mixture incubated at 37C for 30 min.
Polymyxin B treatment of L-RNA: Samples of L-RNA (500/,g/ml) were p!:eincubated at 37C for 30 min with polymyxin B (100#g/ml). The RNA was recovered by ethanol precipitation followed by centrifugation (2000 x g, at 4C for 25min). Samples of LPS (10/zg/ml) were also preincubated with polymyxin B. Treatment of macrophage monolayers with L-RNA or N-RNA: Macrophage monolayers (2 x 107 cells/ well) obtained as described previously were cultured in 2 ml of RPMI medium supplemented with 10% foetal calf serum, 2 mM L-glutamine, RNA from stimulated macrophages induces TNF-cz then removed by washing with RPMI medium. Two of the following stimuli were added to each well: LPS (5 #g/ml); L-RNA (25 #g/ml); L-RNA (25/ig/ml) submitted to enzymatic or polymyxin B pretreatment; N-RNA (25 g/ml); poly A(+)-L-RNA (0.5/.tg/ml), and poly A( )-L-RNA (25/,g/ml). After 90 min incubation at 37C, the supernatants were discarded and the cells washed three times with RPMI, followed by a final 6 h incubation period in RPMI medium (1.5 ml/well).
For TNF or IL-1 assay, the cell-free incubation fluids were subsequently filtered through 0.22 Millipore membranes and stored at -70C until tested. Control wells were incubated with RPMI medium alone.
To investigate the effects of actinomycin-D (10/lg/ml), cycloheximide (10 #g/ml) and dexamethasone (3 #g/ml) on the release of TNF and IL-1 by L-RNA-stimulated macrophage monolayers, the drugs were added to the incubation medium 30 min before the stimuli and throughout the stimulation period. TNF activity assay: The TNF content of the RNA macrophage supernatants was measured using a highly TNF-sensitive cell line, WEHI 164 clone 13, as described elsewhere. 22 Briefly, WEHI cells were seeded in 96-well microplates at a concentration of 4 10 4 cells/well in 50/.tl of supplemented RPMI.
Fifty/tl of serially diluted crude supernatant were added in quadruplicate to the cells. The plates were incubated for 20h at 37C in a 5% CO2-air incubator. Ten #1 of 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) solution (5 mg/ml PBS) were added to each well and the plates incubated for an additional 4h. One hundred #1 of isopropanol containing 0.04 N HCI were then added to each well. Fifteen min later, the degree of cell lysis was evaluated spectrophotometrically (570nm) using an enzyme-linked immunoassay analyser.
Owing to the unavailability of rat TNF, standard curves were obtained with mrTNF-0; the TNF content of the assayed material was calculated by comparison with standard amounts. The results are given as the means of data obtained from at least three different experiments. In some experiments, a rabbit anti-mrTNF antibody was used to demonstrate the specificity of the assay. The results are expressed in international units (IU) of TNF/ml. ILactivity assay: The IL-1 concentrations in supernatants of RNA-stimulated macrophage monolayers were measured using the thymocyte co-mitogenic assay. 23 Briefly, thymocytes obtained from the thymus of C3H/HeJ mice were plated in 96-well microplates at concentrations of 5 105 cells/well, in 100/.tl of supplemented RPMI (see above). One hundred/.tl of serially diluted crude supernatant, with or without concanavalin A (Con A; 0.1-1/ig/well), were added to the incubation medium, and the plates incubated for 72 h at 37C in a 5% CO2-air incubator.
[3H]-Thymidine (0.5/.tCi/well) was then added and the plates incubated for an additional 8 h. IL-1 concentrations were calculated by interpolating the values from standard curves constructed with rhuIL-1/ (0.2-20 IU/well). The results are expressed as IU of IL-1/ml. Secretion of RNA by macrophages stimulated with LPS or cytokines: The detection of RNA in the supernatants of macrophage monolayers was performed using [5-3H]-uridine as a specific precursor of RNA synthesis, according to Kolodny et al. 24 Peritoneal macrophages, obtained from thioglycollate treated rats, were cultured in 12-well tissue culture plates at a concentration of 2 x 10 cells/well, in 2 ml of supplemented RPMI medium. The ceils were allowed to adhere for 1 h at 37C in a 5% CO2-air incubator. The nonadherent cells were removed by washing with RPMI medium, and the macrophage monolayers were incubated with 2 ml of medium containing [5-3H]-uridine (5/lCi/ml) for 2 h. The supernatants were then discarded and the cells washed three times with RPMI medium. To each well 2 ml of the following stimuli were added: LPS (0.1-10#g/ml), rhuIL-lfl (10-10-10-8 M), rhuTNF-0 (10-1-10 -8 M), rhulL-6 (10 -1-10 -8 M) and rhulL-8 (10-1-10 -8 M). Control wells were incubated with RPMI medium alone.
The cell cultures were then incubated for a final 6 h period. After incubation, the supernatants were removed and centrifuged at 10 000 g for 30 min. The RNA was precipitated with 25% TCA at 4C for at least I h. The precipitated radioactive material was filtered on a Millipore system using 0.45/.tm membranes, followed by washing with 5% TCA. Finally, the radioactivity present in the membranes was detected in a liquid scintillation counter were centrifuged at 53 000 x g" for 5 h at 4C and fractionated using a density gradient fractionater. The radioactivity of fractions collected was quantified using a liquid scintillation spectrometer.

Results
The preincubation of macrophage monolayers with cytoplasmic RNA (25 ffg/ml) extracted from LPS (10 g/ml) stimulated macrophages (L-RNA) caused a significant release of TNF and IL-1 in the culture medium as compared with the RNA from non-stimulated macrophages (N-RNA), or with RPMI medium alone (Fig. 1). Such cytokine release was abolished by treatment of the L-RNA samples with RNase but not with polymyxin B. Polymyxin B at the same dose abolished the ability of LPS to stimulate the release of TNF and IL-1. The integrity of our RNA samples was confirmed by SDS-polyacrylamide gel electrophoresis (data not shown).
The release of TNF (Fig. 2) and IL-1 (Fig. 3) by macrophages was induced by both poly A(+) and poly A(--) RNA fractions obtained by chromatography of the L-RNA (total L-RNA) on oligo (dT)--cellulose columns. Although the release of the cytokines was of similar magnitude, the dose of poly A(+) used was 50-fold less than that of poly A(--) or total L-RNA. Figure  stimulated macrophage monolayers was strongly reduced by an antiserum against TNFbut not by the control serum. The effect of TNF-cz antiserum to recombinant murine TNF-0 is also shown in Fig. 2b. The preincubation of macrophage monolayers with the inhibitors of protein synthesis cyclohex-imide (10/.tg/ml) and actinomycin-D (10 #g/ml), or with dexamethasone (3 #g/ml), a general inhibitor of cytokine release, significantly blocked the secretion of TNF induced by L-RNA (Fig. 4a). The release of IL-1 by L-RNA stimulated macrophages was also inhibited by cycloheximide and by dexamethasone (Fig. 4b).
A radioactive material which precipitated with TCA (Fig. 5) and which was sensitive to RNAse (Fig. 6) was detected in the supernatants of macrophage monolayers preincubated with [5-3H] uridine and stimulated with LPS (10/.tg/ml) or rlL-lfl (10 -8 M). These properties suggest that the material was RNA. The release of radioactive RNA was not observed in macrophage monolayers stimulated with rTNF-0 (10-1-10 -8 M) or with LPS in doses of 0.1 or 1 #g/ml, or with IL-lfl in doses 10 -9 or 10-1M (Fig. 5). rIL-6 and rIL-8 (10--10 -8 M) were also ineffective in stimulating the release of RNA (data not shown). The sedimentation profile of the 3H-RNA released by LPS-stimulated or non-stimulated, macrophage monolayers in the sucrose gradient is shown in Fig. 7. The 3H-RNAs secreted under both conditions sediment in the same region of the gradient (4-5 S). However, there was an increase in 3H-RNA secretion by macrophages treated with LPS. Figure 7b shows that the 3H-RNA secreted by LPS stimulated macrophages was completely hydrolysed by KOH pretreatment.  tion profile of the non-radioactive, cytoplasmic RNA is shown in Fig. 7a.
Incubation of the macrophage monolayers with LPS or 1L-lfl in concentrations that stimulated RNA secretion did not alter the viability of the macrophages. Table 1 shows that the LDH activity in the supernatants of macrophages before and after stimulation with LPS (10 #g/ml), IL-1/ (10 -8 M) or TNF-g (10 -8 M) was not different from that observed in the supernatant of macrophages incubated with RPMI medium.

Discussion
The present study shows that macrophage monolayers, when stimulated with L-RNA, release The L-RNA-induced release of cytokines by macrophages is assumed to be RNA specific and not due to LPS contamination, because: (a) this effect is strongly reduced by ribonuclease; (b) N-RNA has no biological effects, thus indicating that the polynucleotide chain by itself is not able to induce cytokine release; and (c) polymyxin B, in a dose that inhibits LPS-induced release of IL-1 and TNFo, does not modify L-RNA activity.
There is much evidence showing that poly A( + )-RNA is the main active biological fraction of exogenous RNA11'20'26'27 despite some findings showing a functional effect for poly A(--)-RNA. 28 In an attempt to identify the fraction of L-RNA active in cytokine release, total cytoplasmic L-RNA samples were fractionated by oligo (dT)-cellulose chromatography and the activity of the poly a(+)and poly A(--)-RNA fractions on IL-1 and TNFrelease were studied. The poly A(+)-RNA (0.5/,g/ml), poly A(--)-RNA (25/g/ml) and total L-RNA (25 #g/ml) had similar effects on the release of the two cytokines.
It is known that poly A(+) containing RNA is the messenger RNA fraction and comprises about 1-2% of the total cellular RNA of eukaryotic cells. 19 In our experiments the dose of poly A(+) that induced the release of TNF-o and IL-1 (0.5/g/ml) corresponds to 2% of the dose of the total L-RNA that also induced cytokine release. This finding is strong evidence that at least part of the L-RNA acts as informational RNA. Corroborating this idea is the fact that cycloheximide, an inhibitor of RNA translation in eukaryotic cells, 29 inhibited the release of TNF-o and IL-1 by L-RNA-stimulated macrophages. The release of these cytokines was also found in the poly A(--)-RNA treated macrophages, suggesting that L-RNA may also act at the transcriptional level. The inhibition of L-RNA activity by actinomycin-D, a classical inhibitor of transcription, confirms this assumption.
It is known that the release of cytokines stimulated by different stimuli is strongly inhibited by glucocorticoids acting at the transcriptional and/or translational levels.  In our experiments dexamethasone blocked the release of IL-1 and TNF from the L-RNA stimulated macrophages which is additional evidence that L-RNA may act at the transcriptional and/or translational levels.
It is now accepted that resident macrophages may act as alarm cells. 33 These cells are highly reactive and, when activated by various stimuli, secrete several mediators including cytokines, which play an important role in triggering and maintaining the defence response. 14'1s In the first part of this paper the authors show that the RNA obtained from LPS-activated macrophages can mediate cytokine release. In the second part of this work, it is shown that these macrophages, when pre-incubated with [5-3H]-uridine and activated with LPS, released a radioactive, TCA-precipitable material into the supernatant which was hydrolysed by RNase and KOH. It is assumed therefore, that these radioactive macromolecules represent RNA which sediments in the 4-5 S region of the sucrose gradient. There are a few reports in the literature 24'34'35 concerning the release by eukaryotic cells of RNA which sediments in the 2.5-5.0 S region of the sucrose gradient. The present observation is the first demonstration that eukaryotic cells release RNA after an exogenous stimulus such as LPS, since previous reports concern spontaneously released RNA.
One interesting finding was that RNA secretion is also induced by an endogenous, pro-inflammatory mediator, IL-lfl, but not by IL-6, IL-8 or TNF-z, suggesting that the release is mediator specific. That RNA secretion may be due to cellular damage was eliminated by the demonstration of cell viability by eosin Y exclusion and by the measurement of lactate dehydrogenase activity in the supernatants of the LPSand cytokinestimulated macrophages. Together, the facts that RNA obtained from activated macrophages can stimulate cytokine release by resting macrophages, and that activated macrophages also release RNA, provide experimental evidence for the concept that RNA may be an intercellular mediator. To obtain further insight concerning the implications of RNA secretion, the biological activity of such secreted RNA is currently under investigation.
In conclusion, our data suggest that resident macrophages, when stimulated by injurious stimuli, in addition to secreting cytokines, are also able to release RNA which may act on the surrounding, resting macrophages, thus stimulating the release of more cytokines. This process may represent amplification of the mechanisms involved in defence responses.